Can Light-Powered Bacteria Recover Heavy Metals from Water? Inside Hokkaido University’s Living Pump

The bacterium is not destroying a metal. It is potentially moving it into a recoverable place.

The dangerous metal does not disappear

A molecule of mercury, cadmium or lead cannot be biodegraded into harmless carbon dioxide the way some organic pollutants can. The challenge in metal-contaminated water is to separate and concentrate ions so they can be safely managed or recovered. In industrial wastewater, that generally means adding chemicals, passing water through separation media, or running energy-intensive processing equipment. What if part of that work could be powered by light and executed by a living cell?

A Hokkaido proposal built around two membrane proteins

At a Japan Science and Technology Agency technology presentation on October 20, 2026, Hokkaido University associate professor Takashi Kikukawa introduced a microbial method for taking up heavy-metal ions from water using light. The Japanese title was 「光で水中の重金属を回収する微生物技術」. According to the JST description, two types of membrane transport protein are introduced into the membrane of Escherichia coli. One uses light energy to generate a hydrogen-ion gradient across the membrane; the other uses that gradient to move metal ions into the cell. By changing operating conditions, the accumulated metals may also be released outside the cell. That final feature distinguishes capture for disposal from a possible reusable recovery cycle.

How light becomes a transport force

The essential concept is a coupling of energy conversion and transport. Protons—hydrogen ions, written H+—are moved across the cell membrane using light. Their resulting electrochemical gradient is a form of stored energy, sometimes compared with a charged battery. A separate metal transporter taps that gradient to drive uptake. These are different tasks carried out by different proteins, not evidence that light itself directly attracts metal ions. The university laboratory studies microbial rhodopsins and ion pumps, a longer-running scientific foundation for this approach. The available presentation does not identify the exact molecular identity of both components, their stoichiometry, or which metal species were tested at commercially relevant concentrations.

Why an E. coli cell is part of the apparatus

Using a living bacterium means the membrane proteins can function in a biological membrane without purifying each protein and rebuilding an artificial membrane system. The trade-off is that cells are not inert filter cartridges. They require suitable conditions and reliable containment; real wastewater may contain salts, acidity, organics and other metal ions that affect growth or transporter performance. Genetically engineered E. coli in an industrial plant would also demand biosafety controls that prevent escape or environmental release. There is no evidence in the JST summary that the method has been approved for discharge into an open river or lake.

Removal is not the same as recovery

Conventional precipitation can convert dissolved metals into solids that then need treatment or disposal. Activated carbon and other adsorbents capture species on surfaces; membrane systems concentrate contaminants in another stream. Each has operating strengths and liabilities. The light-driven approach has a different proposition: a biological collector that takes up ions and could release them under controlled conditions. Whether it ultimately uses fewer reagents or generates less secondary waste remains a hypothesis for engineering verification, not an independently demonstrated full-life-cycle saving.

A research path that predates this week

Kikukawa’s university profile identifies a 2024–2026 research grant for developing E. coli that recovers heavy metals using sunlight and a 2025 project supporting a light-driven metal-recovery nanomachine. The October 20 presentation is therefore a technology-transfer disclosure in a continuing line of research, not proof that scientists built an operational treatment plant during October. His photobiology group has studied how light-sensitive membrane proteins change structure and transport ions. That foundational work matters: designing a recovery system requires understanding which way ions move, how rapidly pumps cycle, and how their functions persist under illumination.

The long history of biological cleanup

Scientists have investigated bacterial binding, uptake and transformation of metals for decades. Many microbes naturally bind metal ions to cell walls or extracellular substances; other systems concentrate them inside cells. Review literature distinguishes passive biosorption from active bioaccumulation. This Hokkaido concept is closer to an engineered active uptake system, because an imported energy-transducing protein supplies a proton gradient. It should not be confused with conventional photosynthesis, nor with bacterial destruction of a toxic element. A recovered metal still requires careful handling.

Where the economic opportunity may lie

JST lists electroplating, mining and metal-refining wastewater as potential treatment settings, plus recycling liquids from batteries and electronic parts. These are meaningful but different markets. In hazardous wastewater the priority may be removal to a regulated concentration; in materials recycling the value may be the purity and saleability of the recovered ions. Extremely dilute metal streams create a special challenge because each liter contains little recoverable value. The team also suggests using the method as a concentration stage before analysis or further processing.

Questions that decide whether it scales

An engineer would first ask which ions the transporter favors and how well it discriminates among competitors. Next come capture capacity per gram of cell biomass, uptake speed, release yield, metal toxicity to the cells, light penetration through turbid liquids, and performance at different pH and salinity. A business case would add reactor size, illumination electricity, cell growth media, sterilization, containment, downstream metal purification and waste management. No validated numerical results for these measures appear in the public presentation summary; we therefore report no removal percentage, cost reduction or treatment throughput.

Japan’s industrial and environmental stakes

Japan has strong manufacturing industries that use metals and a considerable interest in recovering resources from industrial processes and discarded products. Yet public enthusiasm for “green” cleanup can obscure difficult accounting questions: how much energy goes into cultivating microorganisms, supplying light and separating products? Are added chemicals eliminated or merely moved to another stage? Would a process compete with mature treatment systems or work best as a specialized polishing step? Those questions belong at the center of a commercially serious evaluation.

What the October presentation does—and does not—establish

The JST disclosure supports the existence and proposed mechanism of the technology. It states that light-driven metal uptake and conditional metal release are features of the research, and outlines intended applications. It is not a published independently audited demonstration of a municipal-scale deployment. No individual heavy-metal recovery rates, pilot-plant agreements or cost comparisons are specified in the description. Japan.co.jp has not independently interviewed Kikukawa or inspected experimental data. Neither public safety claims nor commercial readiness can be inferred from the presentation alone.

The most intriguing possibility

The arresting idea is not that bacteria will magically clean a lake when sunlight hits the surface. It is that light-responsive molecular machinery can be harnessed to move difficult-to-handle ions into a retrievable biological compartment—and potentially release them again on command. If robust enough for closed industrial systems, that could make biology a component of resource engineering. For now the serious achievement is a plausible scientific architecture and a clear set of difficult questions for the next experiments.

Sources and research documentation